- Keywords: Gene similarity networks, self-organizing maps, convolutional neural networks, multi-omics data integration, graph representation learning, dimensionality reduction
- TL;DR: This paper presents a deep learning model that combines self-organizing maps and convolutional neural networks for representation learning of multi-omics data
- Abstract: One of the main challenges in applying graph convolutional neural networks on gene-interaction data is the lack of understanding of the vector space to which they belong and also the inherent difficulties involved in representing those interactions on a significantly lower dimension, viz Euclidean spaces. The challenge becomes more prevalent when dealing with various types of heterogeneous data. We introduce a systematic, generalized method, called iSOM-GSN, used to transform ``multi-omic'' data with higher dimensions onto a two-dimensional grid. Afterwards, we apply a convolutional neural network to predict disease states of various types. Based on the idea of Kohonen's self-organizing map, we generate a two-dimensional grid for each sample for a given set of genes that represent a gene similarity network. We have tested the model to predict breast and prostate cancer using gene expression, DNA methylation and copy number alteration, yielding prediction accuracies in the 94-98% range for tumor stages of breast cancer and calculated Gleason scores of prostate cancer with just 11 input genes for both cases. The scheme not only outputs nearly perfect classification accuracy, but also provides an enhanced scheme for representation learning, visualization, dimensionality reduction, and interpretation of the results.
- Code: https://gitlab.com/NF2610/isom_gsn